With decreasing defense dollars available to purchase new military aircraft, the inventory of existing aircraft will have to last many more years than originally anticipated. As the avionics computers on these aging aircraft get older, they become more expensive to maintain due to parts obsolescence. In addition, expanding missions and changing requirements lead to growth in the embedded software which, in turn, requires additional processing and memory capacity. Both factors, parts obsolescence and new processing capacity, result in the need to replace the old computer hardware with newer, more capable microprocessor technology. New microprocessors, however, are not compatible with the older computer instruction set architectures. This generally requires the embedded software in these computers to be rewritten. A significant savings-estimated in the billions of dollars-could be achieved in these upgrades if the new computers could execute the old embedded code along with any new code to be added. This paper describes a commercial-off-the-shelf (COTS)-based form, fit, function, and interface (F/sup 3/I) replacement strategy for legacy avionics computers that can reuse existing avionics code "as is" while providing a flexible framework for incremental upgrades and managed change. It is based on a real-time embedded software technology that executes legacy binary code on the latest generation COTS microprocessors. This technology promises performance improvements of 5-10 times that of the legacy avionics computer that it replaces. It also promises a 4/spl times/ decrease in cost and schedule over rewriting the code and provides a "known good" starting point for incremental upgrades of the embedded flight software. Code revalidation cost and risk are minimized since the structure of the embedded code is not changed, allowing the replacement computer to be retested at the "blackbox" level using existing qualification tests.